Prof. T. Q. Bonneij—The Tivt-BiU Conglomerate, 19 



to detect any line of demarcation between the undoubted granitoid 

 beds and the conglomerates. 



2. That in no instance had I been able to find fragments of 

 felsite in the conglomerate, though these abound in the undoubted 

 Cambrian conglomerates ; and yet near Port Dinorwic beds, probably 

 of the same age as those at Twt Hill/ occur close to the felsite, which 

 also in this place appears more probably intrusive in than overlain 

 by the conglomeratic series. 



3. That although the metamorphism of the matrix of the con- 

 glomerate was much less than in the granitoid rock (and in the 

 schist series of Anglesea), yet it seemed greater, especially in the 

 lower and more felspathic band of the Twt Hill conglomerate than 

 is usual in the true Cambrian conglomerate ; vs^hile such beds as the 

 Torridon Sandstone, the Quartzite and the " Upper Gneiss " series of 

 the N.W. Highlands, show how much influence the constitution of 

 a rock has had on the amount of metamorphism produced. 



Thus, although I fully felt the difficulties, being especially 

 troubled by the presence of schist in the conglomerate, I considered 

 the weight of evidence to be in favour of my conclusion. I may say 

 that while each visit to the section made me more confident of the 

 correctness of my view, each revision of the whole question in my 

 study made me more doubtful of it. The same question also has 

 been constantly present to my mind during ray work among meta- 

 morphio rocks in the Alps ; but as conglomerates and even grits 

 appear to be singularly rare in these, the direct evidence which I 

 have obtained has been but slight, and in no way conclusive on 

 either side. 



Dr. Eoberts was led to a conclusion opposite to ours, viz. that the 

 beds were much posterior to the granitoid rock, by the following 

 considerations : — 



1. That he thought he could see a demarcation in the Twt Hill pit 

 between the conglomeratic series and the granitoid rock. 



2. That near Llanerchymedd in Anglesey, a conglomerate, which 

 macroscopically closely resembles the Twt Hill rock, occurs in 

 sequence with fossiliferous Cambrian beds. 



3. That this conglomerate consists of " pebbles of quartz imbedded 

 in a felspathic matrix," which removes the '' only strong d priori 

 argument, viz. the absence of felsite pebbles." 



As regards the validity of this reasoning, I need only point out that 

 it is dangerous to infer the identity of these old rocks without a more 

 thorough examination of their microscopic structure than Dr. Koberts 

 appears to have bestowed upon them ; and that a presumed identity 

 of beds in localities many miles apart is not a very conclusive 

 argument. Indeed, the presence of felsite pebbles in the one and 

 not in the other would be an argument (if for anything) rather 

 against than for such an identity, to which I may add that " a 

 felspathic matrix " — a conveniently vague term — is not to my mind 

 identical with "fragments of felsite." Why did not Dr. Eoberts 



^ I have admitted this point, of which I was at first doubtful. See this Magazine, 

 Vol. VII. Dec. II. p. 300. 



